Optimisation of a geostatistical non-stationary model in history matching

Maria Helena Caeiro, Amilcar Soares, Vasily Demyanov, Michael Andrew Christie

    Research output: Contribution to conferencePaper

    Abstract

    History matching is an important issue in modern reservoir engineering to predict production performance, and the complex non-linear inverse problem one essential topic in reservoir modelling. This work focuses on a history matching problem of non-stationary reservoir models. We propose a novel methodology to link a geostatistical simulation algorithm, namely Direct Sequential Simulation using Local Anisotropy correction with stochastic sampling optimisation methods, specifically the Neighbourhood Approximation Algorithm. This work is based on the approach of simulation of fluvial deltaic depositional structures with locally varying the search
    ellipsoid parameters. The object of study considered was a non-stationary structure (delta front), from Stanford VI – a synthetic petroleum reservoir with fluvial characteristics.
    Original languageEnglish
    Pages1-15
    Number of pages15
    DOIs
    Publication statusPublished - Sep 2011
    EventIAMG 2011 Conference - Salzburg, Austria
    Duration: 5 Sep 20119 Sep 2011

    Conference

    ConferenceIAMG 2011 Conference
    CountryAustria
    CitySalzburg
    Period5/09/119/09/11

    Fingerprint Dive into the research topics of 'Optimisation of a geostatistical non-stationary model in history matching'. Together they form a unique fingerprint.

    Cite this